170 BALFOUR—NEW SPECIES OF PRIMULA. 
circumcinctis, lamina oblonga laevi, petiolo elongato lamina 
multo longiore, fructu et seminibus majoribus recognoscenda. 
Sikkim. Lachen, 11,000-15,000 ft. Hooker. Nos. ro and 
15. June, July 1849. Fl. deep purple like Auricula. In Herb. 
Calc. et Kew. Quite typical. 
Nattong (Gnatong). King’s Collector. No. 4342. 12th July 
1877. Flowers purple; No. 4363. 13th July 1877. Flowers 
dark purple. In Herb. Calc. 
Above Phemgaroo, 12,000 ft. Pantling. May 1885. Flowers 
deep purple. King’s type. (Nos. 46,474, 46,519 in Herb. 
Clarke.) In Herb. Calc. et Kew. 
Hill behind Tangu bungalow, 15,200 ft. Younghusband. 
5th July 1903. In Herb. Calc. et Kew. 
Yatung. Hobson. 1897. In Herb. Kew. 
Sir George King recognised this species, and gave it its name 
on the sheets of the Calcutta Herbarium, but did not publish 
a description. Sir George Watt cites * the name suggesting 
the plant is no more than a variety of P. Kingz, Watt, but 
additional knowledge does not support this view. P. Gammicana, 
‘King is an Eastern Sikkim and Tibetan species having its 
nearest alliance in P. Royle, Balf. fil. et W. W. Sm., a Western 
Sikkim'and Nepal plant. In its type-form marked by King 
it is readily recognised by the long sheathing membranous 
scales below the leaf-rosette, by the long petioles bearing many 
elliptic blades, by the stout pedicels thickening much below the 
calyx in fruit, by the woody stipe of the placenta, and by the 
large seeds. In all these characters it differs from P. Roylet. 
Hooker’s Lachen plants Nos. ro and 15 are also very typical— 
very different from P. Griffithii, Watt, with which Pax joins 
them. P. Gammieana, like P. Roylet, is a purple-flowered 
species formerly referred to P. obtusifolia, Royle, which: is, 
however, a N.W. Himalayan plant of the Nivalis Section, 
whilst P. Gammieana and its ally P. Roylei have characters 
of the Sonchifolia and Petiolaris groups. So far as I know, it 
has not come into cultivation. Only a few undoubted specimens 
of P. Gammieana in herbaria are cited here. But in the 
Calcutta and other herbaria are many specimens resembling 
it, which must be dealt with at another time. Particular 
attention may be directed to specimens from Eumtso La 
and Zemu. Dried material is not very satisfactory for 
the diagnosis of many of these nearly related forms, and 
the limits amongst the forms with which P. Gammieana is 
connected can only be determined by careful study of living 
plants—on the field if possible. Indian botanists in West 
Sikkim may find it worth while to take up the investigation. 
* Watt, On Indian Primulas in Journ. R.H.S., xxix (1904), 300, 314, 319. 
